For good or for ill, it's the truth. The table, background, left/right justify and server push-pull extensions are all at play here. In general they don't keep most non-Netscape users from accessing the information usefully. In fact, we have attempted to check all of our documents cross-client for the best results. In some instances though, coding outside the HTML 2.0 specifications has led to problems.
So why did we choose to do this? While the philosophy of strict conformation to established standards is appropriate for a robust construct, we are interested in exploring some of the limits of existing technologies. We have chosen to model the Proceedings with the widest array of techniques accessible to the largest audience instead of appealing to the lowest common denominator. Obviously this is an optimization problem with no clear best solution.
And then there are the issues of graphic design which are both complicated and simplified by the technology. Working with HTML and the Web can be very frustrating due to client dependencies and the nature of the World Wide Web. There are no guarantees that the document on your screen will look anything like the same information on someone else's screen. Further, there are so many things that one would naturally like to do with the design which are not possible (yet) within the environment. But some decisions have to be made to convey the nature of the materials through the medium.
So when you ask yourself "Why did they do this?", keep in mind that the struggle to represent the materials properly has no 100% satisfying solution. And there are few de facto standards in these early days of network communication. We hope that you are viewing an aesthetically pleasing document which you are finding to be effectively constructed. If you are not, our apologies for not being able to communicate our intentions as we wish we could.
Regardless of your impression, please tell us what you think by going to the Feedback page and leaving your mark on these Proceedings.